Polysubstance Use Among the Homeless In Germany

A nationwide, cross-sectional multicenter study

Background: The number of people experiencing homelessness (PEH) in Germany reached 440 000 in 2024, double the total from 2022. Representative data on substance use among PEH is largely lacking. In this study, we estimate the prevalence of substance use among PEH and identify subgroups at risk of polysubstance use.

Methods: A cross-sectional study of 674 PEH was conducted in four German metropolitan areas in 2021. All PEH were interviewed and provided blood samples in homeless support facilities. Toxicological analysis of serum samples revealed the presence of 22 substances, which were grouped as alcohol, central nervous system (CNS) stimulants, narcotic analgesics, and cannabis. Polysubstance use was defined as the detection of substances belonging to at least two of these groups.

Results: The toxicological analyses revealed that 35% of PEH had no recent substance use (95% confidence interval [31; 39]), while 34% had recently engaged in polysubstance use ([30; 38]). Alcohol was the most prevalent substance (39% [35; 43]), followed by CNS stimulants (30% [27; 34]), cannabis (28% [24; 32]), and narcotic analgesics (18% [15; 21]). Polysubstance use was linked to younger age, prior incarceration, current tobacco use, and geographical location.

Conclusion: In Germany today, the number of PEH is growing, substance availability is widespread, and drug-related deaths are on the rise. It is, therefore, vitally important to continue monitoring the situation and to provide targeted support to those who need it.

Cite this as: Heinrich F, Manthey J, Wulff B, Stallbaum F, Dost K, Graf W, Kowalski V, Brennecke A, Hajek A, König HH, Püschel K, Ondruschka B, Iwersen-Bergmann S: Polysubstance use among the homeless in Germany: A nationwide, cross-sectional multicenter study. Dtsch Arztebl Int 2025; 122: 597–603. DOI: 10.3238/arztebl.m2025.0132

There are currently about 1 300 000 homeless persons (persons experiencing homelessness, PEH) in Europe (1), with around three in four of them suffering from mental disorders (2). Among PEH, disorders caused by alcohol and illegal substance use are the most common psychiatric conditions (2), and illegal substance-related deaths are widespread (3). Data from Denmark and Canada indicate that the standardized all-cause mortality rate among PEH is 3 to 12 times greater than that in the general population (4). Among individuals with previous substance-related hospitalization for the principal cause of death, PEH were seven times more likely to die from illegal substance use, but not alcohol use, than the general population (3).

Although substance use is known to be an integral factor in homelessness and the continuation of homelessness (5), little research into substance consumption in this vulnerable sector of the population has been conducted. Given the lack of a systematic review on the prevalence of substance use among PEH, we reviewed the literature and found only three population-based studies on PEH from different parts of France, Spain, and Canada. In all three countries, PEH were recruited from three major cities, with settings classified as roofless and houseless according to the European Typology on Homelessness and Housing Exclusion (ETHOS) (6). The Canadian study also included PEH from inadequate settings, as defined by ETHOS (e.g., mobile homes or temporary structures). The French study stated the prevalence of tobacco, alcohol, and cannabis use among PEH as 40%, 30%, and 10%, respectively (7), while the Spanish (8) and Canadian (9) studies documented only the prevalence of problematic substance use. In-depth information on the distribution of illicit substance use was not reported in any of these studies. No previous population-based study determined substance use through toxicological analysis.

In Germany, official figures reported 262 000 PEH in 2022, with 178 000 persons registered in emergency accommodation (10). Between 2022 and 2024, the number of persons living in such shelters more than doubled to 440,000 (11). Increasing numbers of PEH have also been observed in other European countries, such as Ireland (1). Furthermore, mortality as a direct consequence of illicit substance use in Germany has almost doubled in the past decade, reaching 2137 deaths in 2024 (12). Polysubstance use, mainly in the form of heroin or other opioids mixed with other substances, was a common cause of death (13), suggesting that this is a widespread and hazardous practice.

The current study aims to provide detailed information about substance and polysubstance use among PEH in Germany in the period July to September 2021. Using a population-based approach and relying on self-reporting and toxicological data, this study provides a comprehensive overview of the use of specific substances and their combination and explores risk factors for polysubstance use.

Methods

Study design

This was a national cross-sectional study in four German metropolitan areas: Hamburg, Frankfurt am Main (including Mainz and Wiesbaden), Leipzig (including Halle an der Saale), and Munich (including Augsburg). The National Survey on the Psychiatric and Somatic Health of Homeless Individuals (NAPSHI) was conducted between July and September 2021 (14). This study was designed in adherence to the Declaration of Helsinki and approved by the ethics committee of the Hamburg Medical Association (application number PV7333). Reporting followed the STROBE guidelines (Strengthening the Reporting of Observational Studies in Epidemiology). Details of the recruitment process can be found in eSupplement Material 1. Data on the homeless support landscape were obtained through an online survey of homeless support facilities (eSupplement Material 2).

Primary outcome: Polysubstance use

The primary outcome, the prevalence of polysubstance use, was examined by means of toxicological analyses (see eSupplement Material 3 for details). Polysubstance use was defined as the use of substances from at least two groups, comprising alcohol, central nervous system (CNS) stimulants, narcotic analgesics, and cannabis Details of each see substance group examined can be found in eSupplement Table 1. Single substances were identified as positive when testing for the substance or its metabolite yielded a positive result. Substance groups were identified as positive when at least one substance or metabolite was positive and negative when all substances and metabolites were negative. For some PEH, individual substance group measurements were missing when insufficient samples were available to measure the remaining substances. These missing measurements were assumed missing at random (n = 37). Polysubstance use was determined only for complete cases.

Table 1

Description of the sample of PEH in Germany by sex (n = 674); 25 PEH did not report their sex

Secondary outcome: Single substance use

As secondary outcome, the prevalence of use of the following substances was determined from blood samples and self-reports: alcohol (blood alcohol and desialotransferrin [carbohydrate-deficient transferrin, CDT]), amphetamine, methamphetamine, cocaine, MDMA (3,4-methylenedioxymethylamphetamine, “ecstasy”), opiates, medicinal opioids, substitution opioids, and cannabis. The questionnaire used to assess self-reported substance use is shown in eSupplement Material 4.

Exposure

Self-reported sociodemographic characteristics, somatic and psychiatric health status, and geography constitute exposures of interest. No effect modification was assumed.

Bias

Selection bias was avoided by the multicenter recruitment in homeless support facilities. To minimize barriers to participation in the study, questionnaire-based interviews were offered so that illiterate persons could also take part. The questionnaires were available in several languages, and a translator was called upon when necessary. Self-report bias was addressed by performing objective toxicological measurements. Multiple imputation was used to handle missing data.

Statistical analysis

The sample for analysis included all PEH eligible for study enrolment. The methods used for statistical analysis can be found in eSupplement Material 5.

Graphical illustration

GraphPad Prism (Version 10.2.3 [347]; GraphPad Software, CA, USA) and Adobe Illustrator 2024 (Version 28.5; Adobe, CA, USA) were used for graphical illustration.

Results

In total, 674 PEH were enrolled from four metropolitan areas in Germany: 213 from Hamburg, 108 from Leipzig, 154 from Frankfurt, and 199 from Munich. The mean age was 44 years (standard deviation [SD] 12), with an 18% proportion of women (n = 118). The PEH were recruited from settings classified as rooflessness (sleeping rough, without any form of shelter; 57%, n = 353) and houselessness (sleeping in temporary shelters or institutions; 43%, n = 262). Table 1, eSupplement Table 2, and eSupplement Table 3 contain further information on housing status and sample characteristics.

Table 2

Positive toxicological tests for substances and substance groups in PEH in Germany by sex (n = 674); 25 PEH did not report their sex

Table 3

Self-reported substance use by sex (n = 674); 25 PEH did not report their sex

Primary outcome: Polysubstance use

The primary outcome was evaluated among PEH with complete toxicological data (n = 583). No substantial differences from PEH with incomplete toxicological analysis were observed (eSupplement Table 3). In one third of the PEH, no recent substance use was detected (n = 201; 34.5% [95% confidence interval [30.6; 38.5]). Conversely, two thirds of the PEH (n = 382; 65.6% [61.5; 69.4]) had recently used one of the four substance groups examined. In every third PEH (n = 198; 34.0% [30.1; 38.0]), polysubstance use was identified (≥ 2 substance groups; male 36% vs female 28%). Among PEH who had recently used any substance, a median of two substance groups were consumed (interquartile range [IQR] 1 to 4). As shown in Figure 1, the most common two-way combinations of substance groups were alcohol with CNS stimulants, alcohol with cannabis, CNS stimulants with cannabis, and CNS stimulants with narcotic analgesics. Distinct patterns of substance group use can be found in eSupplement Table 4.

Figure 1

An arc plot illustrating co-use patterns across four substance classes

After controlling for other sociodemographic and health variables, polysubstance use was significantly associated with age, history of incarceration, active tobacco use, and geography (Figure 2). If all PEH had reported having been in prison, the proportion of polysubstance use would have been 39%. In contrast, if none of the PEH had been in prison, the proportion of polysubstance use in this sample would have been 14%. Results from the complete case analysis are shown in eSupplement Table 5. Only about half of the surveyed sites allow PEH to use substances on their premises (illegal substance use: 13%; alcohol use: 23%).

Figure 2

Odds ratios and 95% confidence intervals from pooled multivariable logistic regression, after substantive-model compatible multiple imputations with 50 imputation data sets and 1000 iterations between the imputations

Secondary outcome: Single substance use (blood-based)

Overall, alcohol was the single most frequently used substance, followed by cannabis, cocaine, amphetamines and methadone (Table 2). The eSupplement Figure shows the serum concentrations of the respective substances and metabolites among PEH testing positive for substance use. Blood alcohol concentrations (median 1.2‰, IQR 0.5–2.0) and CDT concentrations (median 4.70%, IQR 3.00–8.60) were very high.

Secondary outcome: Single substance use (self-reported)

About 40% of PEH reported consuming more than seven glasses of alcohol per week. More than 40% exceeded sex-specific thresholds for heavy episodic alcohol consumption on a daily to weekly basis (Table 3). Every sixth PEH stated they used substances intravenously, with the majority of PEH having done so for many years. Every second PEH stated they had used illegal substances in the past year. The substances mostly concerned were cannabis, cocaine, and street opiates.

Discussion

Among 674 PEH from four metropolitan areas in Germany, two thirds had recently used alcohol, cannabis, or illegal substances. One third of the PEH were found to have recently used substances from at least two groups, most commonly alcohol together with CNS stimulants or alcohol and cannabis. Among PEH with documented substance use, polysubstance use appears to be a common practice.

This is the first population-based study to estimate the prevalence of substance use by means of toxicological measurements in a large, heterogeneous sample of PEH. At first sight, the high prevalence of substance use among PEH is not surprising and corresponds to previous studies (2). However, our findings add a critical dimension: a tripartition of PEH into abstainers, single-substance users, and polysubstance users. Considerable heterogeneity of substance use can be observed among PEH: one third did not test positive for any substance, while another third tested positive for two or more substances. Very high blood levels of alcohol and CDT, together with a high proportion of PEH engaging in heavy episodic drinking on a daily or weekly basis, indicate severe chronic alcohol use. The high rates of cardiovascular and liver problems observed in this population (14) may be caused by regular alcohol intake over a long period (15). Nevertheless, it needs to be stressed that a considerable number of PEH do not use any substances. In fact, the proportion of abstainers may be similar to that in the general population or even larger (16). In 2021, three out of ten adults in Germany reported that they had not consumed alcohol in the previous month. There are no data on the use of other substances remains unclear (16). With both abstainers and heavy users represented among PEH, shared spaces may be difficult to establish. In most of the facilities surveyed, use of alcohol or illegal substances is not allowed, creating a barrier to heavy users but potentially creating a safe environment for abstainers. By acknowledging the differential needs of PEH, services can be tailored to meet the specific requirements of different subgroups. Essentially, this means housing programs that presuppose recovery from substance use problems (17); such programs have shown promising health impacts (18).

A number of studies have revealed that incarceration constitutes a major, partly avoidable risk factor for continued homelessness (19) and health consequences (20). The study presented here shows, in addition, that previous imprisonment is a strong predictor of current polysubstance use. In our sample, 54% of PEH reported having been imprisoned at some point in their lives. Imprisonment can result from criminal offenses or non-payment of administrative fines. In Germany, a substitute custodial sentence can be imposed for non-payment of administrative fines (e.g., as a result of not paying transport fares). Such sentences disproportionately affects socioeconomically disadvantaged persons, including PEH and those with substance use problems (21).

Several limitations of this study need to be acknowledged. First, it is difficult to establish true representativeness. As there is no register of either facilities or PEH, we cannot rule out possible selection biases due to non-response. This is further limited by the lack of response rates in our and many other studies on PEH. Many studies on PEH underrepresent women (22), and we may not be an exception in this regard (the proportion of women in our study is 18%; official estimates range around 35%) (10). Yet, considering various forms of homelessness in four regions, our sample of PEH is believed to be approximately representative of PEH in Germany in 2021. Second, serum samples were employed for toxicological analyses using highly sensitive measurement techniques. Blood measurements generally limit the detection window to about 24 to 48 hours before enrolment, contingent on substance and frequency of use. This constraint precludes the direct extrapolation of our findings to long-term patterns of substance use.

In conclusion, our findings show that polysubstance use was identified in around one third of PEH—in addition to a multitude of other environmental and social risk factors. Within this already vulnerable population, those engaging in polysubstance use may represent the subgroup at greatest risk, given their elevated rates of incarceration. Against the backdrop of a growing PEH population (10), wide availability of legal and illegal substances (23), and escalating rates of illicit substance overdose deaths in Germany (12), continued monitoring and targeted support of those in need appear especially warranted.

Acknowledgments

We thank the staff at the facilities for the homeless and the participating scientists for their support.

Funding

FH and FB were funded by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG; grant number 493624519). This study was financially supported by the Volkswagen Foundation (project number 99269). The local branch of the German Red Cross provided a vehicle free of charge during the entire data collection period.

Conflict of interest statement
KP, FH, and BO received support for manuscript compilation from the Volkswagen Foundation. JM’s institution received funding from the German Federal Ministry of Health JM himself has received consultation fees from the AOK and the WHO, as well as fees for presentations on the topic of substance use from the BAS, the DHS, the Friedrich Ebert Foundation, the Voluntary Welfare League of Sachsen–Anhalt, and the Therapy Store (Therapieladen e. V.). Travel costs were paid by the charitable foundation of Hanover Medical School. The remaining authors declare that no conflict of interest exists.

Manuscript received on 5 January 2025, revised version accepted on 14 July 2025

Corresponding author
Fabian Heinrich, MD, MSc

References

1. Feantsa: Homelessness in Europe – The State of Play 2024 [cited 2025 17.06.]. www.feantsa.org/public/user/Activities/events/2024/9th_overview/EN_Chap/1.pdf (last accessed on.7 August 2025).

2. Gutwinski S, Schreiter S, Deutscher K, Fazel S: The prevalence of mental disorders among homeless people in high-income countries: An updated systematic review and meta-regression analysis. PLoS Med 2021; 18: e1003750.

3. Morrison DS: Homelessness as an independent risk factor for mortality: Results from a retrospective cohort study. Int J Epidemiol 2009; 38: 877–83.

4. Aldridge RW, Story A, Hwang SW, et al.: Morbidity and mortality in homeless individuals, prisoners, sex workers, and individuals with substance use disorders in high-income countries: A systematic review and meta-analysis. Lancet 2018; 391: 241–50.

5. Thompson RG, Wall MM, Greenstein E, Grant BF, Hasin DS: Substance-use disorders and poverty as prospective predictors of first-time homelessness in the United States. Am J Public Health 2013; 103 (Suppl 2): S282–S8.

6. Feantsa, Abbé Pierre Foundation: Eight Overview of housing exclusion in Europe 2023. www.feantsa.org/public/user/Resources/reports/2023/OVERVIEW/Rapport_EN.pdf (last accessed on.7 September 2025).

7. Scarlett H, Melchior M, Davisse-Paturet C, et al.: Substance use among residents of homeless shelters during the COVID-19 pandemic: Findings from France. Int J Public Health 2022: 67:1604684.

8. Parés-Bayerri A, Calvo F, Font-Mayolas S, Panadero S, Vázquez JJ: Differences in drug use among persons experiencing homelessness according to gender and nationality. Int J Environ Res Public Health 2023; 20: 4007.

9. Zhang L, Norena M, Gadermann A, et al.: Concurrent disorders and health care utilization among homeless and vulnerably housed persons in Canada. J Dual Diagn 2018; 14: 21–31.

10. Bundesministeriums für Arbeit und Soziales: Ausmaß und Struktur von Wohnungslosigkeit 2022. www.bmas.de/SharedDocs/Downloads/DE/Soziale-Sicherung/wohnungslosenbericht-2022.html (last accessed on 7 September 2025).

11. Statistisches Bundesamt: Ende Januar 2024 rund 439 500 untergebrachte wohnungslose Personen in Deutschland. Pressemitteilung Nr. 282 vom 15.7.2024.

12. Der Beauftragte der Bundesregierung für Sucht- und Drogenfragen: Zahl der Drogentoten in Deutschland weiter angestiegen. Pressemitteilung vom 29.05.2024.

13. Der Beauftragte der Bundesregierung für Sucht- und Drogenfragen: Rauschgift-Todesfälle 2023. 2024.www.bmas.de/SharedDocs/Downloads/DE/Soziale-Sicherung/wohnungslosenbericht-2022.html (last accessed on 1 September 2025).

14. Bertram F, Hajek A, Dost K, et al.: The mental and physical health of the homeless—evidence from the National Survey on Psychiatric and Somatic Health of Homeless Individuals (the NAPSHI study). Dtsch Arztebl Int 2022; 119: 861–8.

15. Rehm J, Gmel Sr GE, Gmel G, et al.: The relationship between different dimensions of alcohol use and the burden of disease—an update. Addiction 2017; 112: 968–1001.

16. Rauschert C, Möckl J, Seitz NN, Wilms N, Olderbak S, Kraus L: The use of psychoactive substances in Germany—findings from the Epidemiological Survey of Substance Abuse 2021. Dtsch Arztebl Int 2022; 119: 527–34.

17. O‘Shaughnessy BR, Mayock P, Kakar A: The recovery experiences of homeless service users with substance use disorder: A systematic review and qualitative meta-synthesis. Int J Drug Policy 2024: 130: 104528.

18. Onapa H, Sharpley CF, Bitsika V, et al.: The physical and mental health effects of housing homeless people: A systematic review. Health Soc Care Community 2022; 30: 448–68.

19. Nilsson SF, Nordentoft M, Fazel S, Laursen TM: Risk of homelessness after prison release and recidivism in Denmark: A nationwide, register-based cohort study. Lancet Public Health 2023; 8: e756–e65.

20. Reimer S, Pearce N, Marek A, Heslin K, Moreno AP: The impact of incarceration on health and health care utilization: A system perspective. J Health Care Poor Underserved 2021; 32: 1403–14.

21. Bögelein N, Glaubitz C, Neumann M, Kamieth J: Bestandsaufnahme der Ersatzfreiheitsstrafe in Mecklenburg-Vorpommern. De Gruyter Brill 2019; 102: 282–96.

22. Women‘s Homelessness: European Evidence Review [press release]. Feantsa 2021; 01.04.2021.

23. Deutsche Hauptstelle für Suchtfragen e. V. DHS Jahrbuch Sucht 2025. Lengerich: Pabst; 2025. www.dhs.de/fileadmin/user_upload/pdf/Jahrbuch_Sucht/JBSucht2025_komplett_WEB.pdf (last accessed on 7 August 2025)